Diesel Cheaper than Petrol

Forum for general chat, news, blogs, humour, jokes etc.

Postby revian » Sat Jul 25, 2015 9:09 am


Mr Cholmondeley-Warner wrote:
Silk wrote:You make it sound as if old cars regularly made it to 15 or 20 years old and could be keep running with nothing more than a bit of know-how and a 13 mil spanner.

1/2" AF, please! None of your new-fangled metric bollox thanks :mrgreen:

Absolutely! I've just given away my taps and dies though...
Wirral
revian
 
Posts: 509
Joined: Fri Aug 02, 2013 6:37 pm

Postby Silk » Sat Jul 25, 2015 11:15 am


akirk wrote:having just bought a 43 year old MGB GT I know that virtually any garage in the country can strip it down and rebuild it...


But you're looking at it from the point of view of an enthusiast. You may not want to hear it, but for most people, an old MG is an over-rated pile of junk, that no one in their right mind would want to own in preference to a modern car. Modern cars don't need stripping down and rebuilding every couple of months and are quite capable of going 'round the clock twice with not much more than regular oil changes and a few bushes and filters. Modern cars usually end up in the scrapyard because the latest repair bill exceeds the cost of replacement, not because they can't be fixed because some computer chip has failed.

Although, at first glance, you would think that new cars with all their gizmos, turbos and emissions stuff would be a disaster waiting to happen, but I doubt there is any evidence to back this up. Time will almost certainly show that the current crop of new cars are more reliable than those that went before and there's no reason to believe this won't continue to be the case.
Silk
 
Posts: 1033
Joined: Sun Apr 13, 2008 2:03 pm

Postby akirk » Sat Jul 25, 2015 10:11 pm


Silk wrote:
akirk wrote:having just bought a 43 year old MGB GT I know that virtually any garage in the country can strip it down and rebuild it...


But you're looking at it from the point of view of an enthusiast. You may not want to hear it, but for most people, an old MG is an over-rated pile of junk, that no one in their right mind would want to own in preference to a modern car. Modern cars don't need stripping down and rebuilding every couple of months and are quite capable of going 'round the clock twice with not much more than regular oil changes and a few bushes and filters. Modern cars usually end up in the scrapyard because the latest repair bill exceeds the cost of replacement, not because they can't be fixed because some computer chip has failed.

Although, at first glance, you would think that new cars with all their gizmos, turbos and emissions stuff would be a disaster waiting to happen, but I doubt there is any evidence to back this up. Time will almost certainly show that the current crop of new cars are more reliable than those that went before and there's no reason to believe this won't continue to be the case.


really?!

I have been in the computer industry for over 25 years and it is abundantly clear that this is not the case - technology of this nature tends to have issues from new - let alone when old, if it is not sorted out when created, what makes you think that it will be easy to fix when 20-40 years old?

I have a brand new Skoda Octavia Scout on the drive, it has a setting in the audio system to auto switch off the screen - you set it, turn off the ignition, it forgets the setting. You press pause on the music - now switch off the ignition, some times when you switch back on it will still be paused, sometimes not. This is brand new technology and it doesn't work - if skoda can't get it right new, how do you believe it will be fixed in 20 years time?

I can get every part of my MG either new, or restored and could put it back on the road exactly as it was 43 years ago when built - I am fairly confident that will not be the case for current cars in 40 years time. I will be amazed if the same percentage of modern cars is on the road as for old cars... in that timescale...

as for having to strip down and rebuild MGs every few months - have you ever owned / driven one? I have run one as my only car before and had no issues and no particular maintenance required...

Alasdair
akirk
 
Posts: 668
Joined: Thu Oct 30, 2014 11:07 am
Location: Cotswolds

Postby Silk » Sun Jul 26, 2015 11:39 am


StressedDave wrote:The problem isn't with the technology itself per se, but the fact that they employ idiots (note expletive deleted) on minimum pay (for the industry) to write the code that drives it. And further the issue is focus groups driven by the marketeers who end up with a list of specifications as long as your arm that have to be implemented for no cost and in no reasonable timescale. I bet their test cases are crap as well, otherwise this would be ironed out.


I'm not sure your attack on software developers is helpful - I sense anger management issues. If you can get a job writing code, even if it's bad code, then you're far from being an idiot.

The reality is, manufacturers are in fierce competition to get the latest gizmo out there before someone else does and, it doesn't matter how much you pay the developers, they can only work so fast. It's the same throughout the tech industry. Beta testing takes time and people won't wait. These days, beta testing is carried out by the end user with the supplier providing a constant stream of patches. The mobile phone industry is an example of this, where end users would rather have the latest software with the bugs than last year's without. Even Microsoft puts out pre-release software to anyone who can't wait for the final release, even if this means they end up being unpaid beta testers. People just have to have the latest and they get what they deserve.

The radio described above will have been manufactured by a third party, probably in China, that supplies lots of brands. The car only supplies basic data such as, "is the ignition on?", "are the lights on?" etc. Because of the higher costs, the core parts of the car will have a longer development cycle, so there will be fewer "bugs", hopefully.

In short, there's no reason to believe that new cars are not more reliable than ever.
Silk
 
Posts: 1033
Joined: Sun Apr 13, 2008 2:03 pm

Postby revian » Sun Jul 26, 2015 2:45 pm


StressedDave wrote: The problem isn't with the technology itself per se, but the fact that they employ idiots (note expletive deleted) on minimum pay (for the industry) to write the code that drives it. And further the issue is focus groups driven by the marketeers who end up with a list of specifications as long as your arm that have to be implemented for no cost and in no reasonable timescale. I bet their test cases are crap as well, otherwise this would be ironed out.


I'm not sure it's that simple any more in separating 'the technology' from the 'code that drives it'. Any software which is above the aeomeba stage might have so many combinations that running through all the possibilities is easier said than done.

It's been true of all types of technology that the list of enhancements looked for grows but alongside a driving down of costs. 'Of course it can be made cheaper and better' is the driving mantra....'. In the deep and dark past I worked in a cost reduction team for BT Development Dept of second generation TXE4 exchanges.

In the motor industry of 20 years ago (guessing it's not much different now) you didn't say to a manufacturer 'this is what it'll cost you for x.' They say; 'This is what we will pay you. Take it or leave it'.

As to building things which 'only' have to respond to the signals given to it and send out other signals in response....that's how the control centres of the latest nuclear submarines are designed and built. Hope they have factored the radio in properly...
Ian
Wirral
revian
 
Posts: 509
Joined: Fri Aug 02, 2013 6:37 pm

Postby trashbat » Sun Jul 26, 2015 3:02 pm


That's not quite how it works - yes, the number of combinations tends towards infinity, but (a) only some are valuably different and (b) software isn't designed, implemented and tested as a single unit, any more than a house or a car is a single unit of design.

I don't think it's a case of peanuts and monkeys either, more a case of poor system design and risk analysis. And no, nuclear submarine software or anything like it is designed in a very different way to most of the consumer-grade stuff in your car. It still won't be perfect.
Rob - IAM F1RST, Alfa Romeo 156 JTS
trashbat
 
Posts: 764
Joined: Wed Jun 22, 2011 11:11 pm
Location: Hampshire

Postby revian » Sun Jul 26, 2015 4:35 pm


trashbat wrote:That's not quite how it works - yes, the number of combinations tends towards infinity, but (a) only some are valuably different and (b) software isn't designed, implemented and tested as a single unit, any more than a house or a car is a single unit of design.

Thank you.... That's what I was trying to stumble towards before I had a post-lunch doze! How easy is it to be sure that some 'valuably different' outcomes haven't been missed? Do you work outwards from the code possibities or back to check the code from worst case scenarios. ... I'm not sure I'm being clear and may need to sleep again :wink:
trash at wrote:I don't think it's a case of peanuts and monkeys either, more a case of poor system design and risk analysis. And no, nuclear submarine software or anything like it is designed in a very different way to most of the consumer-grade stuff in your car. It still won't be perfect.

...as nothing ever really is.

I'm very aware I'm more than a tad rusty on this. I had HNC qualifications in computing (analogue and digital) before they existed in such common ways and certainly before I owned a calculator. System costs were Iinputted over a telex link from High Holborn to Dollis Hill.... Results came back by post!

Ian
Wirral
revian
 
Posts: 509
Joined: Fri Aug 02, 2013 6:37 pm

Postby Pyrolol » Sun Jul 26, 2015 4:48 pm


If you ask me, accusing software of significantly impacting the long term maintainability of modern cars sounds wrong.

After a few years in the field, most of the software issues that are significant enough to cause real problems (i.e. ones bad enough to justify fixes) should have been ironed out. There'll always be some left, but fundamentally the highest risk time for software issues bricking your device is the first few months after release - not 20 years later. Once it's working well enough for a particular person's use pattern, they have a good chance of not hitting further problems.

Contrast that with hardware, which (after the early deaths) gets less reliable over time as it wears out.
Pyrolol
 
Posts: 26
Joined: Wed Sep 11, 2013 11:19 pm
Location: Enfield

Postby akirk » Sun Jul 26, 2015 4:52 pm


Silk wrote:
StressedDave wrote:The problem isn't with the technology itself per se, but the fact that they employ idiots (note expletive deleted) on minimum pay (for the industry) to write the code that drives it. And further the issue is focus groups driven by the marketeers who end up with a list of specifications as long as your arm that have to be implemented for no cost and in no reasonable timescale. I bet their test cases are crap as well, otherwise this would be ironed out.


I'm not sure your attack on software developers is helpful - I sense anger management issues. If you can get a job writing code, even if it's bad code, then you're far from being an idiot.

The reality is, manufacturers are in fierce competition to get the latest gizmo out there before someone else does and, it doesn't matter how much you pay the developers, they can only work so fast. It's the same throughout the tech industry. Beta testing takes time and people won't wait. These days, beta testing is carried out by the end user with the supplier providing a constant stream of patches. The mobile phone industry is an example of this, where end users would rather have the latest software with the bugs than last year's without. Even Microsoft puts out pre-release software to anyone who can't wait for the final release, even if this means they end up being unpaid beta testers. People just have to have the latest and they get what they deserve.

The radio described above will have been manufactured by a third party, probably in China, that supplies lots of brands. The car only supplies basic data such as, "is the ignition on?", "are the lights on?" etc. Because of the higher costs, the core parts of the car will have a longer development cycle, so there will be fewer "bugs", hopefully.

In short, there's no reason to believe that new cars are not more reliable than ever.


Sadly SD is right - how you define and critique them is irrelevant, the people employed to do these jobs are not doing them correctly and it is not complicated. They can't have decent test scenarios - it is not difficult: turn unit on - press each button and go through each setting - what happens, is that as designed. check clicking settings on and off - do both do as they should, and does it stick. Probably take no more than 1-2 hours for the Skoda's system, could that not be factored in to development time? It took me about 20 minutes to discover that a setting doesn't save - that is utterly basic testing...

as for how things are tested - depends on the design, if it is digital in its intent (i.e. there are fixed options - such sas the setting being on or off) then it is simple to test, each scenario tested and if there is impact elsewhere, tested in combination... - if it is more analogue in intent, (e.g. Saatnav entry text) then you might have test scenarios or scripts - or indeed, like the Skoda you code to make it more defined, so in the Skoda while you can enter free form text it is always filtering against its list of known entries and only allows those - easier to test, but infuriating to use when it doesn't have the name of the road from which I enter my drive!

speed of putting out new features & software bugs are not entirely what I was reflecting on in terms of reliability and long term longevity - it is the latter which will be more of an issue - even if code works perfectly now and the car is 100% reliable now - that doesn't give it longevity, the complexity of systems and the ability to rebuild / re-engineer them affects that - and simply put computers are likely to be more difficult to rebuild / recode / restore in 20-40 years time than a bit of steel with rust in it...

Alasdair
akirk
 
Posts: 668
Joined: Thu Oct 30, 2014 11:07 am
Location: Cotswolds

Postby Pontoneer » Sun Jul 26, 2015 5:10 pm


Silk wrote:
akirk wrote:having just bought a 43 year old MGB GT I know that virtually any garage in the country can strip it down and rebuild it...


But you're looking at it from the point of view of an enthusiast. You may not want to hear it, but for most people, an old MG is an over-rated pile of junk, that no one in their right mind would want to own in preference to a modern car. Modern cars don't need stripping down and rebuilding every couple of months and are quite capable of going 'round the clock twice with not much more than regular oil changes and a few bushes and filters. Modern cars usually end up in the scrapyard because the latest repair bill exceeds the cost of replacement, not because they can't be fixed because some computer chip has failed.

Although, at first glance, you would think that new cars with all their gizmos, turbos and emissions stuff would be a disaster waiting to happen, but I doubt there is any evidence to back this up. Time will almost certainly show that the current crop of new cars are more reliable than those that went before and there's no reason to believe this won't continue to be the case.


Sorry , but I can't agree .

Having grown up with Mercedes-Benz since childhood : I remember my dad buying his first one in 1964 and we've had every generation in the family since then ; my first car was a 220/8 from 1970 handed down from my dad when I passed my driving test in 1975 , and I've gone through every generation since then - those cars from the 1950's/60's/70's/80's were reliable when new and continue to be now , with proper maintenance .

I made the mistake of buying a 2003 S203 C270CDI at a cost of £4000 : BIG mistake ; I'd had one Diesel car before , which was a W124 E250 diesel , and it was a good car . This later car was nothing but trouble : in the two years I had it all five diesel injectors failed , not a cheap matter at close to £300 each , the electric door locking mechanism in the drivers door failed ( £70 ) , the tailgate wiper failed (£200 ) , what would have been a £12.50 ball joint in the front suspension on an earlier model failed , needing the whole suspension link arm replaced at £170 , latterly the swirl flaps in the inlet manifold failed with a likely parts cost of £400 - all the prices I have mentioned are parts only , labour would have been more - never in my 40 years of motoring had I had so much trouble from one car , or so much cost to repair - the idea had been to buy a cheap to run diesel car - it did return circa 50 mpg , but the repair costs over my 2 years ownership more than wiped out the fuel savings . If that was my experience with the car at 10 years old I dread to think what it would be like at 20 , 30 , 40 or more years old .

I replaced it with a 1993 190E , bought for £250 , I've replaced the engine due to a failed cylinder head gasket ( £200 used engine from a club member was the cheapest option ) and plan to overhaul the brakes , but otherwise has been trouble free in 18 months , and anything that goes wrong I can fix myself . I also swapped the C270 for a 1993 300SL-24 which I knew would need work - so far I've replaced the steering wheel as the original had been removed leaving SRS and horn inoperable ! I fitted a new battery and am about to have a new exhaust system fitted on Tuesday due to the original 22 year old one being rusted out ( can't complain at that ) before the MOT ; otherwise that car has been reliable over the last 3 months and I'm confident I can fix anything that might go wrong as it's just a W124 under the skin .

All of my older cars are/were simple mechanical things and can be kept running by any competent home mechanic ; the newer stuff is often beyond even the main dealers who , if anyone can , ought to be able to diagnose and fix them - if they can't do so now , what hope is there in 20-40 years time ? Since new cars often need individual parts being coded into their systems , even a trip to the scrappy ( as I can currently do with my existing cars ) won't be an option in the future , so once the run of the mill six years after discontinuation parts availability expires , these cars may as well be scrap metal .
Pontoneer
 
Posts: 96
Joined: Sat Jul 16, 2011 4:03 pm

Postby triquet » Sun Jul 26, 2015 5:42 pm


We go over to France once a month and there is a very good reason for having a diesel barge, particularly with the current state of the Euro ... :mrgreen:
Jim
Offshore Engineer and Master of Music
triquet
 
Posts: 265
Joined: Mon Jul 07, 2014 12:51 pm

Postby trashbat » Sun Jul 26, 2015 6:03 pm


revian wrote:Thank you.... That's what I was trying to stumble towards before I had a post-lunch doze! How easy is it to be sure that some 'valuably different' outcomes haven't been missed? Do you work outwards from the code possibities or back to check the code from worst case scenarios. ... I'm not sure I'm being clear and may need to sleep again :wink:

It's a whole different discussion, but...

- you design everything in small, closed units or modules with their externally-available behaviours exposed through a small number of interfaces

- if you're really doing it right, you write the tests before you implement anything (test driven development) and then implement the software specifically to pass them

- you test the internal behaviours within those units

- you also poke the interfaces within the intended set of parameters, and then some unintended ones to make sure it's handled or rejected appropriately

- if security is a concern, you then do what's called penetration testing, where you take one of these units, then with and without looking at the code, try to exploit it

Generally you work in a layered approach from the code up in terms of these units and modules, and by the end you do some high level system testing where you throw stuff at the finished product to see what you can do to that too.
Rob - IAM F1RST, Alfa Romeo 156 JTS
trashbat
 
Posts: 764
Joined: Wed Jun 22, 2011 11:11 pm
Location: Hampshire

Postby revian » Mon Jul 27, 2015 9:19 am


Thanks Trashbat... I'm not so far out as I feared.

As Pontoneer wrote... A major problem is the re-coding for replacement parts ruling self help largely out.... (Oops -split infinitive :wink: ) Plus the dealer being less and less of a mechanic and more of a swap it out/in shop. Few people understand whole systems anymore... To be fair it's a harder task now.

Ian
Wirral
revian
 
Posts: 509
Joined: Fri Aug 02, 2013 6:37 pm

Previous

Return to General Car Chat Forum

Who is online

Users browsing this forum: No registered users and 5 guests